Interesting Theresa Senft article entitled
Against Reputation
Interestingly, although there have been criticisms of specific implementations of reputation management systems (i.e., the existence of plain dumb reviewers on Amazon) I have yet to see a full-on argument regarding what I see as the biggest problem with reputation itself: its reliance on a spherical mode of relationality, as in the phrase, "sphere of influence." Spheres are ways of delimiting space, and with it, people and ideas. Just as the our understanding of the public sphere turns on how we define public (and lock out those who don't fit in), to claim a sphere of influence, one must first declare in advance of the interaction "these are the people/ideas currently influencing me, and these are the ones who do not."
This is timely, with Seb's reference to
LJ trying to unpack "the overloaded concept of friend", which I think is going to be a disaster. There are some places where, when technology is made _more_ flexible, it gets notably _less_ usable;
we cannot ever render human relations with complete explicitness (Paging Dr. Weinberger to the white courtesy telephone...).
Taking the label 'Friend' on LiveJournal (whose primary virtue is that it is obviously inadequate, so people don't read as much into it), and turning it into something multi-variate and hard to use, and which will still be inadequate but now confusingly so, since after all shouldn't we oughta be able to say exactly what we mean? (to which the answer is of course "No", and the history of computer science's encounters with real people has largely been the history of misunderstanding that constraint), but, as I say, taking the simple term Friend and thinking it will be a good idea to sub-divide and sub-divide and sub-divide til
Jason Kottke is hiring someone to manage _just_ his LiveJournal profile is like trying to open a can to get out just one worm -- it seems like a good idea right up to the moment you open the can.
1. Bill Seitz on January 30, 2004 9:01 PM writes...
I think the fundamental problem is that there's no function to provide a context to the labelling.
A label is like a model or prediction. But without anything to predict the label has little point. Or you could think of it as a design approach without a problem to test fitness against.
It's like a lot of pointless attempts at defining ontologies/taxonomies for no clearly intended benefit.
Permalink to Comment2. Jon on January 31, 2004 12:06 AM writes...
Dividing into endless minutiae.
Software that helps us connect with each other, one way or another, will be with us from now on.
It'll be just like on dry land - some people you connect with you'll like and trust pretty fast, and maybe or probably do stuff with - and maybe you won't really know why you like and trust them.
Othere who on paper you should like and trust you won't get along with and won't do anything with.
Some people who you clearly don't like and don't trust you'll end up doing something with (business or otherwise) because you need or want something from them badly enough, or because they need or want something badly enough. Ever had an offer you just can't refuse, or slept with someone you really didn't want to.
We're humans, fer crissake. If Orkut doesn't cut it, some other software will come along, and another after that. It'll keep getting better unti it's just good enough for us to keep on being humans but with applications that help us connect to other people we may or may not like, may or may not trust, and may or may not do things with. Just like in *real* life, with all the applicable caveats. And we'll get fooled and surprised just as often, and in just as many different ways.
Permalink to Comment3. Dan Steinbock on January 31, 2004 7:09 AM writes...
The ghettoization of Livejournal will certainly spell cultureshock for a community raised from birth under (approximately) consistent laws, but I don't see how this implies that partitioning of social networks in general equates with segregation. Granted there is something satisfying about the de-categorised social relations we've grown in networkspace. Maybe it appeals to our desire for simplicity, or an ideal of social unity. Or perhaps it just fits best with our intuitive cognitive categories: self & Other. I sympathize with any digital community that dares to alter the (arbitrary) implementation under which it bloomed.
However, when the number of our electronic "friends" far exceeds our natural cognitive limits, the social network loses its salience. Transient interactions in netspace often have a greater permanence than in physical space simply because they are preserved in the digital record. Massive friend lists on Orkut and Friendster cease to represent anything other than the persuasive abilities of individuals to accrue links for the sake of links. This may be marginally interesting by itself as an artificial phenomenon, but in practice these networks tend towards incoherence.
If we are serious about putting social networks to work then we cannot forever postpone the partition issue. The practical usefulness of the social network is greatly reduced when many types of social relationship are conflated into homogeneity. It makes sense to distinguish my business connections from my romantic ones, my musical collaborators from my athletic collaborators. A link given a context is a link that means something; it becomes a relationship upon which we can begin to build transitive inferences. If I share musical tastes with person A and A shares musical tastes with B, it's probable I'd like the same music as B. This is information that can be leveraged. This is a social network that actually does something besides passively map out generic relationality.
That said, it is problematic to predefine a rigid global set of possible relationship types. If there are too few you restrict the ability of individuals to untangle highly related but highly populated clusters (as a music enthusiast I might want to distinguish friends who like jazz from those who like indie-rock). Too many and the network becomes sparse and disconnected--dividing into endless minutiae. Perhaps you should let every individual have their own categorization scheme, applicable only to their own list-of-friends (I distinguish jazz-heads & indie-rockers while you lump them under music-lovers; Dan likes many categories, Terri likes one). Then everyone would be free to impose their own personal idea of what a "friend" is. But then what good is meaning if it is not shared meaning? Again we lose the ability to make useful inferences.
As yet there is no trivial answer to the problem of the bloated friends-list, but we're working on it. At the very least, I prefer my social networks to have non-uniform connection strengths so I can sort the list by degree of connectedness. But inevitably, as the population of meaningful acquaintances outgrows the limits of manageability, we will need context.
Superficially, it's a user interface issue. But Terri's deep concern serves as a cautionary tale for would-be designers of social software: even simple UI decisions can have catastrophic effect on the community culture of the people who use it.
Permalink to Comment4. Jon Husband on January 31, 2004 9:48 AM writes...
"Superficially, it’s a user interface issue. But Terri’s deep concern serves as a cautionary tale for would-be designers of social software: even simple UI decisions can have catastrophic effect on the community culture of the people who use it"
Indeed ! Just like the design of an urban space will have critical effects on the types of neighbourhoods, and the rhythyms and shapes of life that take place there, and the livability - not to mention the extent to which people get to know their neighbours or not. See people like Jane Jacobs or Anne Galloway for wisdom in this domain.
It's probably wise to develop specific niche SN software for specific purposes, that incorporate the "just good enough" aspects of whatever socializing means in that niche's context - that's why there are Rotary Clubs, and book clubs, and Boards of Trade, and neighbourhood associations - and within those there are usually a few people you meet you get close to, and all the others you smile at and shake hands when the group has its more formal get togethers and then forget about till the next time.
Maybe women should design business networking social software and men should design relationship building software, and then we might all evove faster ?
Permalink to Comment5. Teller Coates on January 31, 2004 7:12 PM writes...
I'm reminded of Ebay's feedback system. One black mark and all hell breaks loose. A truly honest system would cause our warts to stick out.
Permalink to Comment